Power testing with an oscilloscope - where does the hard clip come from?

Discussion in 'Amp Tech Center' started by peteb, Nov 4, 2019.

  1. peteb

    peteb Friend of Leo's

    Posts:
    4,105
    Joined:
    Apr 25, 2003
    Location:
    Cascadia
    The OT? The power tube(s)? The preamp? All of the above?



    I just started power testing this week. Tried it on a couple different amps.



    What I found. There is no real technique to finding the max power signal. Any reasonable signal level, frequency, or amp settings will do. The wave is nicely shaped until it reaches its limit or ceiling and then it just conforms to that level and does the hard level cut.



    Where does the hard clip usually occur in these kind of tests?



    First I was measuring a champ. Got it up to 5 Watts on the scope and figure it must be the power tube getting close to its 6W limit.


    Then I was looking at a 12 watt Princeton. It gets up to 10-11 W on the scope. Is that the tubes reaching their limit (12 W each) or is it the undersized output transformer reaching its limit? The power tubes are biased cool, but they could be biased just as hot as a hot biased 22W deluxe reverb.



    Back to the champ. I was able to get distinctly different breakup on either end of the wave. The hard straight clip must be saturation and the angled clip must be cut off. Saturation happened right before cutoff, meaning the bias could be cooled a little. How is it certain the hard clip is saturation and not cutoff? Most likely because the Princeton has the hard clip on both ends of the wave and that is saturation. Not sure why the cutoff is not visible at all on the Princeton.



    So I’m pretty sure the hard clip is power tube saturation. To make sure I scope the power tube itself. I have measured signal on the grid and plate before with a DMM so I had some idea what to expect. I was going to scope the plate of the power tube, expecting a couple hundred volts but then I noticed the voltage scale setting on the O-scope only goes up to 20 V/cm so I ruled that out. So I scoped the grid. I had the idea that when you scope different places on the amp it always looks the same. This is probably not true because signal tracing would not work. What did the grid look like on the scope? It looked weird. There were three signals instead of one and they were at an angle, but it still did the hard clip at high volume.





    This is the real question that necessitated this post.



    If I saw a hard clip on the grid of the power tube, does that mean that the clip started in the pre amp?
     
    Last edited: Nov 4, 2019
  2. 2 Headed Goat

    2 Headed Goat Tele-Afflicted

    Posts:
    1,851
    Joined:
    Jul 28, 2012
    Location:
    age 10
    The valves (tubes) are what's amplifying the signal, the output transformer is just matching the impedance to the load (speaker).

    The driver tube/phase inverter is amplifying the signal to the point where the power tubes can maximise the gain factor.

    A clipped signal on the grid of an output tube would be caused by the previous gain stage.
     
    Last edited: Nov 5, 2019
  3. LightningPhil

    LightningPhil TDPRI Member

    Posts:
    60
    Joined:
    Dec 1, 2018
    Location:
    Nottingham, UK
    Assume you’re using a sine wave from a signal generator. Try different frequencies. Gain will be quite different throughout the guitars range and vary rather a lot between amps.

    It’s not the power limit of the tubes that causes clipping, it’s running into voltage limits. I run some of my amps a little low on voltage to prolong the life of the tubes, my hearing and relationship with neighbours while still getting cranked sounds. About 85 to 90% if anyone’s curious. Others have ended up about 5% over voltage after some rectifier and tank cap mods. They’re loud! And when messing about with a Variac, it’s been fun playing with voltages between 20% and 125% of what they should be. Lower the rail voltage, the earlier it clips.

    At some point, too higher voltage, combined with lots of current through the valves will over heat them. But consider the power rating of the tubes much more of a guide than a rule.
     
  4. peteb

    peteb Friend of Leo's

    Posts:
    4,105
    Joined:
    Apr 25, 2003
    Location:
    Cascadia

    Thanks Goat,

    I agree but I disagree. The grid voltage is from the previous stage, but previously I had developed the notion, from testing, that the signal shows itself thru out. I, I can’t explain.



    This is why I disagree. The one circuit element that sets the power level for the amp, above all others, is the power tubes. The power tubes is where the power level is set.
     
    2 Headed Goat likes this.
  5. peteb

    peteb Friend of Leo's

    Posts:
    4,105
    Joined:
    Apr 25, 2003
    Location:
    Cascadia
    Thank you Phil,


    I agree. 1K offers a little more power than 440, but not by much. The tone settings also might make a little difference.




    Getting to the questions of what causes the straight across clipping, I’m leaning towards tube saturation, pre amp or power amp. It just kinda makes sense. There is a level where the tube will pass no more power. And it is not tweekable, the tube is the tube.
     
    Last edited: Nov 5, 2019
  6. Peegoo

    Peegoo Tele-Meister

    Age:
    59
    Posts:
    174
    Joined:
    Oct 11, 2019
    Location:
    USA
    It's caused by the tubes' limitations; beyond a certain input voltage level, they cannot take a smooth-sine-wave signal, amplify it, and make a smooth sine wave signal. This can also happen in the preamp (too hot a signal fed to the preamp stage), which is what an overdrive or boost does.
     
  7. Bendyha

    Bendyha Friend of Leo's Silver Supporter

    Posts:
    2,082
    Joined:
    Mar 26, 2014
    Location:
    Northern Germany
    In reference to the above, I refer back to this thread, maximum wattage 2 6V6, where you once again refute the facts with bogus claims based on misconceptions:twisted:

    A circuit, is just that, a loop containing interactive elements, each limited not only by its own surrounding component settings, its own characteristics and tolerances, but by everything from the pickup winding and magnets, through to the speaker dopeing and cab resonances.

    As to where the clipping starts (its compounding development) well, that is what a o'scope is for, finding and comparing the input to the various stages of progression through the circuit...where it starts, is where you find it starts.
     
    LudwigvonBirk and Wally like this.
  8. LightningPhil

    LightningPhil TDPRI Member

    Posts:
    60
    Joined:
    Dec 1, 2018
    Location:
    Nottingham, UK
    You could try putting a large, surplus cpu heat-sink on your output tubes to allow more capacity before clipping.

    Ok, that’s silly. But tubes power throughput is limited by keeping its temperature low enough not to cook. The hotter they are, the shorter the life. So if the class A biasing is high, then much power will be burnt doing nothing and the tube will always be toasty. But if more class B biased as in a push pull amp, then they will idle cooler. That allows much, much higher power transients. So a class B amp can be configured to run higher output for short periods of time. Useful for an abundance of clean headroom before clipping sets in.

    In essence, the power rating is guidance. An EL84 might say it’s power rating is 12W (from which about 10W output is available just as clipping sets in). But for long life, running it cooler at 10W input would be good and reduce output power a bit. Some run them at 14W input for a bit more oomph. Their life still seems OK. 20W for brief periods is fun too. But as the required voltage goes up, and the guts get hotter, there’s more noise. Having a box of them helps feel brave while carrying out such silly experiments.

    Totally. At the cabinet resonance the driver has an impedance many times that of its DC rating, which is reflected back through the output transformer and changes the behaviour of the output tubes.
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.


  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.